37563e59bd
This is a fix for a similar bug that remained, as in #235. `Content::Image#to_h` returned `mime_type` (snake_case), but the MCP spec requires `mimeType` (camelCase) for the MIME type field in image content.
33 lines
967 B
Ruby
33 lines
967 B
Ruby
# frozen_string_literal: true
|
|
|
|
require "test_helper"
|
|
|
|
module MCP
|
|
module Content
|
|
class ImageTest < ActiveSupport::TestCase
|
|
test "#to_h returns mimeType (camelCase) per MCP spec" do
|
|
image = Image.new("base64data", "image/png")
|
|
result = image.to_h
|
|
|
|
assert_equal "image/png", result[:mimeType]
|
|
refute result.key?(:mime_type), "Expected camelCase mimeType, got snake_case mime_type"
|
|
assert_equal "image", result[:type]
|
|
assert_equal "base64data", result[:data]
|
|
end
|
|
|
|
test "#to_h with annotations" do
|
|
image = Image.new("base64data", "image/png", annotations: { role: "thumbnail" })
|
|
result = image.to_h
|
|
|
|
assert_equal({ role: "thumbnail" }, result[:annotations])
|
|
end
|
|
|
|
test "#to_h without annotations omits the key" do
|
|
image = Image.new("base64data", "image/png")
|
|
result = image.to_h
|
|
|
|
refute result.key?(:annotations)
|
|
end
|
|
end
|
|
end
|
|
end
|